Exhibit 99.3

Title: Body mass index and weight are predictors of pregnancy in a Phase 3 multicenter contraceptive efficacy study of AG200-15, a low-dose combination hormonal contraceptive patch

Authors: Thomas D. Kimble(a), Alfred Poindexter(b), Kurt Barnhart(c), Paula M. Castaño(d), Beatrice A. Chen(e), Joseph A. Chiodo III(f), Elizabeth I.O. Garner(f)

(a)Eastern Virginia Medical School, 601 Colley Ave, Norfolk, VA; (b)Advances in Health, Inc., 7515 South Main St Suite 360, Houston, TX; (c)University of Pennsylvania Medical Center, 3701 Market St Suite 810, Philadelphia, PA; (d) Columbia University Irving Medical Center, New York, NY; (e)University of Pittsburgh/Magee-Womens Research Institute, 300 Halket St, Pittsburgh, PA; (f)Agile Therapeutics, 101 Poor Farm Rd, Princeton, NJ.

Objectives: To evaluate predictors of contraceptive failure (i.e., pregnancy) for an investigational low-dose weekly transdermal contraceptive containing levonorgestrel and ethinyl estradiol (AG200-15).

Methods: The SECURE (Study to Evaluate Contraceptive Use, Reliability and Effectiveness) study was a single-arm, open-label, multicenter Phase 3 trial that enrolled 2032 sexually active females >18 years old at risk for pregnancy. A new patch is applied weekly for three weeks and no patch is worn during week four. There were no enrollment restrictions on weight or body mass index (BMI). We performed univariate logistic regression analyses to identify variables predictive of contraceptive failure and to inform two stepwise regression analyses (SRA1 and SRA2) for the primary efficacy analysis population (women <35 years old).

Results: Weight and BMI variables were highly correlated in univariate analysis; we therefore performed SRA1 for BMI >30 kg/m2 and <30 kg/m2 and SRA2 for weight deciles.  Variables included were age, race, ethnicity, smoking status, alcohol use, education level, prior use of hormonal contraception, number of prior pregnancies, late/missed patch application, unscheduled patch removal, patch adhesion, compliance, and obesity (SRA1) or weight (SRA2). SRA1 found the odds of contraceptive failure were 1.96 (1.19, 3.23) times greater for obese versus non-obese subjects.  SRA2 found for every increase of one decile in weight, the odds of contraceptive failure were multiplied by 1.15 (1.05, 1.25).

Conclusions: Previous studies have reported that obesity affects the efficacy of combined hormonal contraceptives.  Data from the SECURE study supports this hypothesis for a low-dose combination hormonal contraceptive patch.